CN110263680A - Image processing method, device and system and storage medium - Google Patents
Image processing method, device and system and storage medium Download PDFInfo
- Publication number
- CN110263680A CN110263680A CN201910477796.2A CN201910477796A CN110263680A CN 110263680 A CN110263680 A CN 110263680A CN 201910477796 A CN201910477796 A CN 201910477796A CN 110263680 A CN110263680 A CN 110263680A
- Authority
- CN
- China
- Prior art keywords
- face
- facial image
- image
- optimal
- quality
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/94—Hardware or software architectures specially adapted for image or video understanding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/161—Detection; Localisation; Normalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/172—Classification, e.g. identification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30168—Image quality inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30196—Human being; Person
- G06T2207/30201—Face
Abstract
The embodiment of the present invention provides image processing method, device and system and storage medium.Method includes: real-time image acquisition sequence；Face datection is carried out to each image in image sequence in real time, facial image corresponding to each face is obtained in the case where detecting face, and analyze the face quality of each facial image；Face recognition module, which is sent, by the earliest facial image that face quality in the facial image for belonging to same target object reaches preset requirement earliest carries out recognition of face；Cache that face is best in quality in the facial image for belonging to target object and face quality is more than the optimal facial image of earliest facial image in optimal frame queue；At the scheduled time, check the image buffer storage situation of optimal frame queue, if the face for being cached with the optimal facial image and special object that belong to special object in optimal frame queue not yet identifies, face recognition module is sent by the optimal facial image for belonging to special object and carries out recognition of face.Useless calculating can be effectively reduced.
Description
Technical field
The present invention relates to field of face identification, relates more specifically to a kind of image processing method, device and system and deposit
Storage media.
Background technique
In field of face identification, the facial image of image collecting device (such as camera) acquisition someone can use, with
The facial image that can use acquisition afterwards carries out recognition of face.It appears in camera view in a people to leaving camera view
In the process, camera can collect the facial image of many this persons, if each facial image is sent to recognition of face mould
If block carries out recognition of face, huge computational load can be brought for face identification system, be easy to cause that equipment heating is hot, meter
The decline of calculation ability, the problems such as power consumption is more, cruise duration is short, it is most important that will increase useless calculating because a people once at
It just no longer needs to that identification is repeated several times after function identification.
Summary of the invention
The present invention is proposed in view of the above problem.The present invention provides a kind of image processing methods, device and system
And storage medium.
According to an aspect of the present invention, a kind of image processing method is provided.This method comprises: real-time image acquisition sequence；
Face datection is carried out to each image in image sequence in real time, is obtained in the case where detecting face corresponding to each face
Facial image, and analyze the face quality of each facial image；By face matter in the facial image for belonging to same target object
The earliest facial image that amount reaches preset requirement earliest is sent to face recognition module and carries out recognition of face；In optimal frame queue
Caching belongs to that face in the facial image of target object is best in quality and face quality is more than the optimal face of earliest facial image
Image；At the scheduled time, check the image buffer storage situation of optimal frame queue, if be cached in optimal frame queue belong to it is specific right
The optimal facial image of the elephant and face of special object not yet identifies, the then optimal facial image that will belong to special object are sent
Recognition of face is carried out to face recognition module.
Illustratively, face quality in the facial image for belonging to same target object is reached into the earliest of preset requirement earliest
It includes: that real-time judge belongs in the facial image of target object often that facial image, which is sent to face recognition module and carries out recognition of face,
Whether the face quality of a facial image reaches preset requirement, until earliest facial image occurs；And by earliest facial image
It is sent to face recognition module and carries out recognition of face.
Illustratively, face quality in the facial image for belonging to same target object is reached into the earliest of preset requirement earliest
It includes: that real-time judge belongs in the facial image of target object often that facial image, which is sent to face recognition module and carries out recognition of face,
Whether the face quality of a facial image reaches preset requirement；Current face's image in the facial image for belonging to target object
Face quality reach preset requirement in the case where, judgement belongs in the facial image of target object with the presence or absence of any first forefathers
Face image is already sent to face recognition module, is earliest facial image if there is no then determining current face's image；It will most
Early facial image is sent to face recognition module and carries out recognition of face.
Illustratively, cache in the facial image for belonging to target object that face is best in quality and face in optimal frame queue
Quality is more than that the optimal facial image of earliest facial image includes: in the facial image for belonging to target object, by first man
Face quality is more than that the facial image of the face quality of earliest facial image is buffered in optimal frame queue as optimal facial image；
When acquisition belongs to current face's image of target object every time, by the face quality of current face's image and optimal facial image
Face quality compares；In the case where the face quality of current face's image is more than the face quality of optimal facial image,
The optimal facial image cached in optimal frame queue is updated to current face's image.
Illustratively, target object and special object are same targets, at the scheduled time, check the image of optimal frame queue
Caching situation include: in the case where carrying out recognition of face failure based on earliest facial image, check in optimal frame queue whether
It is cached with the facial image for belonging to target object, wherein predetermined instant is to carry out recognition of face failure based on earliest facial image
Any time later.
Illustratively, at the scheduled time, the image buffer storage situation for checking optimal frame queue includes: every predetermined period, inspection
Look into the facial image for whether being cached in optimal frame queue and belonging to any object, wherein special object is that facial image is buffered in
Any object in optimal frame queue.
Illustratively, after at the scheduled time, checking the image buffer storage situation of optimal frame queue, method further include: such as
The face that the optimal facial image and special object that belong to special object are cached in the optimal frame queue of fruit has identified, then will belong to
It is removed in the optimal facial image of special object from optimal frame queue and/or stores the optimal facial image for belonging to special object
In memory.
Illustratively, when carrying out Face datection to each image in image sequence in real time, method further include: for dividing
Not Wei Yu two faces in two different images of image sequence, the position in the Face datection result based on two faces
Whether diversity judgement two faces between information belong to same target object.
According to a further aspect of the invention, a kind of image processing apparatus is provided, comprising: module is obtained, for obtaining in real time
Image sequence；Detection and analysis module is detecting people for carrying out Face datection to each image in image sequence in real time
Facial image corresponding to each face is obtained in the case where face, and analyzes the face quality of each facial image；Sending module,
Earliest facial image for face quality in the facial image for belonging to same target object to be reached to preset requirement earliest is sent
Recognition of face is carried out to face recognition module；Cache module, for caching the face for belonging to target object in optimal frame queue
Face is best in quality in image and face quality is more than the optimal facial image of earliest facial image；Module is checked, for pre-
Timing is carved, and is checked the image buffer storage situation of optimal frame queue, is belonged to the optimal of special object if be cached in optimal frame queue
The face of facial image and special object not yet identifies, then sends face for the optimal facial image for belonging to special object and know
Other module carries out recognition of face.
According to a further aspect of the invention, a kind of image processing system, including processor and memory are provided, wherein institute
It states and is stored with computer program instructions in memory, for executing when the computer program instructions are run by the processor
State image processing method.
According to a further aspect of the invention, a kind of storage medium is provided, stores program instruction on said storage,
Described program instruction is at runtime for executing above-mentioned image processing method.
Image processing method, device, system and storage medium according to an embodiment of the present invention, use earliest facial image with
And the recognition of face strategy that optimal facial image combines, this strategy can effectively reduce useless calculating, identify face as early as possible, together
When can improve as far as possible recognition of face go out probability.
Detailed description of the invention
The embodiment of the present invention is described in more detail in conjunction with the accompanying drawings, the above and other purposes of the present invention,
Feature and advantage will be apparent.Attached drawing is used to provide to further understand the embodiment of the present invention, and constitutes explanation
A part of book, is used to explain the present invention together with the embodiment of the present invention, is not construed as limiting the invention.In the accompanying drawings,
Identical reference label typically represents same parts or step.
Fig. 1 shows showing for the exemplary electronic device for realizing image processing method according to an embodiment of the present invention and device
Meaning property block diagram；
Fig. 2 shows the schematic flow charts of image processing method according to an embodiment of the invention；
Fig. 3 shows the schematic diagram of an exemplary image processing flow according to the present invention；
Fig. 4 shows the schematic block diagram of image processing apparatus according to an embodiment of the invention；And
Fig. 5 shows the schematic block diagram of image processing system according to an embodiment of the invention.
Specific embodiment
In order to enable the object, technical solutions and advantages of the present invention become apparent, root is described in detail below with reference to accompanying drawings
According to example embodiments of the present invention.Obviously, described embodiment is only a part of the embodiments of the present invention, rather than this hair
Bright whole embodiments, it should be appreciated that the present invention is not limited by example embodiment described herein.Based on described in the present invention
The embodiment of the present invention, those skilled in the art's obtained all other embodiment in the case where not making the creative labor
It should all fall under the scope of the present invention.
The embodiment of the invention provides a kind of image processing method, device and system and storage mediums.According to the present invention
The image processing method of embodiment, the recognition of face strategy combined using earliest facial image and optimal facial image are this
Strategy can effectively reduce useless calculating, identify face as early as possible, while can improve the probability that recognition of face goes out as far as possible.According to this
The image processing method and device of inventive embodiments can be applied to any required field for carrying out recognition of face, such as security protection door
The fields such as taboo, e-commerce, banking.
Firstly, describing the example for realizing image processing method according to an embodiment of the present invention and device referring to Fig.1
Electronic equipment 100.
As shown in Figure 1, electronic equipment 100 includes one or more processors 102, one or more storage devices 104.It can
Selection of land, electronic equipment 100 can also include input unit 106, output device 108 and image collecting device 110, these groups
Part passes through the interconnection of bindiny mechanism's (not shown) of bus system 112 and/or other forms.It should be noted that electronics shown in FIG. 1 is set
Standby 100 component and structure be it is illustrative, and not restrictive, as needed, the electronic equipment also can have it
His component and structure.
The processor 102 can use digital signal processor (DSP), field programmable gate array (FPGA), can compile
At least one of journey logic array (PLA), microprocessor example, in hardware realizes that the processor 102 can be centre
It manages unit (CPU), image processor (GPU), dedicated integrated circuit (ASIC) or there is data-handling capacity and/or instruction
The combination of one or more of the processing unit of other forms of executive capability, and can control the electronic equipment 100
In other components to execute desired function.
The storage device 104 may include one or more computer program products, and the computer program product can
To include various forms of computer readable storage mediums, such as volatile memory and/or nonvolatile memory.It is described easy
The property lost memory for example may include random access memory (RAM) and/or cache memory (cache) etc..It is described non-
Volatile memory for example may include read-only memory (ROM), hard disk, flash memory etc..In the computer readable storage medium
On can store one or more computer program instructions, processor 102 can run described program instruction, to realize hereafter institute
The client functionality (realized by processor) in the embodiment of the present invention stated and/or other desired functions.In the meter
Can also store various application programs and various data in calculation machine readable storage medium storing program for executing, for example, the application program use and/or
The various data etc. generated.
The input unit 106 can be the device that user is used to input instruction, and may include keyboard, mouse, wheat
One or more of gram wind and touch screen etc..
The output device 108 can export various information (such as image and/or sound) to external (such as user), and
It and may include one or more of display, loudspeaker etc..Optionally, the input unit 106 and the output device
108 can integrate together, be realized using same interactive device (such as touch screen).
Described image acquisition device 110 can acquire image, and acquired image is stored in the storage device
For the use of other components in 104.Image collecting device 110 can be the camera etc. in individual camera or mobile terminal.
It should be appreciated that image collecting device 110 is only example, electronic equipment 100 can not include image collecting device 110.This
In the case of, it can use other device acquisition images with Image Acquisition ability, and the image of acquisition transmission electron is set
Standby 100.
Illustratively, the exemplary electronic device for realizing image processing method according to an embodiment of the present invention and device can
To be realized in the equipment of personal computer or remote server etc..
In the following, image processing method according to an embodiment of the present invention will be described with reference to Fig. 2.Fig. 2 shows according to the present invention one
The schematic flow chart of the image processing method 200 of a embodiment.As shown in Fig. 2, image processing method 200 includes the following steps
S210, S220, S230, S240 and S250.
In step S210, real-time image acquisition sequence.
Image sequence may include at least one image.Each image in image sequence can be still image, can also
To be the video frame in one section of video, in the later case, described image sequence is this section of video.It is every in image sequence
A image can be the original image that image acquisition device arrives, and be also possible to pre-process original image (all in full
Word, normalization, smooth etc.) after the image that obtains.
Image sequence can come from external equipment, by external equipment be transmitted to electronic equipment 100 carry out image procossing or
Carry out image procossing and subsequent recognition of face.In addition, image sequence can also be adopted by electronic equipment 100 for face
Collection obtains.For example, electronic equipment 100, which can use image collecting device 110 (such as independent camera), acquires at least one
Image, to obtain image sequence.Acquired image sequence can be transmitted to processor 102 by image collecting device 110, by
Device 102 is managed to carry out image procossing or carry out image procossing and subsequent recognition of face.
Step S220 carries out Face datection to each image in image sequence in real time, in the case where detecting face
Facial image corresponding to each face is obtained, and analyzes the face quality of each facial image.
Image in image sequence can be to be obtained successively, in real time, one present image of every acquisition, then to the current figure
Facial image corresponding to each of present image face is obtained as progress Face datection and in the case where detecting face,
And analyze the face quality of each facial image.
Face datection can be carried out using method for detecting human face that is any existing or being likely to occur in the future.Art technology
Personnel are understood that the implementation of various method for detecting human face, do not repeat herein.
By Face datection, the Face datection result of each face included in each image can be obtained.It is exemplary
Ground, Face datection result may include the location information for being used to indicate corresponding face position.For example, bounding box can be used
(bounding box, may be simply referred to as bbox) represents the position where face.Illustratively, bounding box can be rectangle frame.Show
Example property, the location information of bounding box corresponds to the location information of face, can be indicated with four numerical value.For example, any people
The location information of face can be shown with following numerical tabular: the upper left corner abscissa x of bounding box corresponding to the face, the vertical seat in the upper left corner
Mark the height h of y, the width w of bounding box, bounding box.In another example the location information of any face can be corresponding to the face
The coordinate representation on four vertex of bounding box.
Limitation is only exemplary rather than with the representation that bounding box represents face location.For example, can be with employment on the face pre-
If the key point and/or facial contour line of number represent face location, correspondingly, the location information of any face may include institute
State the profile point coordinate on the coordinate and/or facial contour line of the key point of preset number.The key point of preset number can wrap
Include the central point of such as face.
After carrying out Face datection for present image, it can be separately included in the image from being obtained in present image
The facial image of each face, each facial image can exclude as far as possible other faces and other classes mainly comprising corresponding face
The interference (such as building etc.) of type.In one example, each facial image can be the correspondence image from image sequence
The image block comprising face corresponding to the facial image extracted, or scaling etc. is carried out to the image block extracted
The image obtained after processing.For example, it is assumed that an image includes two faces, then it can extract and separately include from the image
Two image blocks of two faces.The image block extracted can be used for the operation such as subsequent recognition of face.It is extracting comprising appointing
When the image block of one face, the image block of same size can be extracted for each face, also can according to need extraction not
With size image block and then by scaling etc. processing normalization.In another example, each facial image can be figure
As the correspondence image in sequence, or the image that correspondence image is scaled, is obtained after the processing such as noise reduction.The latter
The case where example especially suitable for each image in image sequence at most only includes a face, may not need in this case
For the individual image block of each face extraction.
For facial image, it is understood that there may be many indexes are used to measure the quality of face quality.For example, can basis
Human face posture (can be indicated with facial angle, i.e. the angle that deflects to certain directions of face), image in facial image is fuzzy
One of indexs such as degree, face occlusion state, light conditions a variety of measure face quality.
Specifically, for example, if face side face angle or pitch angle are more than angle threshold, it may be considered that face quality
It is unqualified, it is believed that its requirement for being unable to satisfy recognition of face accuracy., whereas if face side face angle or pitch angle do not surpass
Over-angle threshold value, it may be considered that face is up-to-standard.In another example if the fog-level of facial image is more than Fuzzy Threshold,
It is also assumed that face is off quality., whereas if the fog-level of facial image is no more than Fuzzy Threshold, it may be considered that
Face is up-to-standard.In another example if some key positions (for example, eyes and/or mouth) in face are blocked, then it is assumed that
Face is off quality., whereas if the key position in face is not blocked, then it is assumed that face is up-to-standard.In another example such as
The illumination brightness of fruit facial image is lower than luminance threshold, then it is assumed that face is off quality., whereas if the illumination of facial image
Brightness is not less than luminance threshold, then it is assumed that face is up-to-standard.For another example many indexes can be comprehensively considered, such as in face figure
The fog-level of picture is more than Fuzzy Threshold or brightness of image lower than in the case where luminance threshold, it is believed that face quality does not conform to
Lattice.Conversely, in the case where the fog-level of facial image is no more than Fuzzy Threshold and brightness of image is not less than luminance threshold,
It is considered that face is up-to-standard.It will be understood by those skilled in the art that the synthesis of These parameters is exemplary, this hair
Bright to be not limited thereto, those skilled in the art can also carry out a variety of synthesis to above-mentioned each index according to actual needs.
It illustratively, can be by index involved in face quality, such as human face posture, figure for each facial image
One of picture fog-level, face occlusion state, light conditions etc. are a variety of, combine and calculate a quality score.
In this case, the face quality analysis results of each facial image may include the quality score of the facial image.Quality
The height of scoring can indicate the quality of face quality.The quality score can be compared with preset quality threshold, to sentence
Whether disconnected face quality is qualified.Quality threshold can be any suitable threshold value, can be set as needed, and the present invention is not right
This is limited.
It can be using face mass analysis method analysis face quality that is any existing or being likely to occur in the future.It is exemplary
And not restrictive, it can use the face quality of convolutional neural networks analysis facial image.For example, above-mentioned difference can be directed to
Different convolutional neural networks are respectively trained in index (for example, human face posture, image fog-level and face occlusion state), often
A convolutional neural networks export the scoring of corresponding index respectively, can finally integrate and obtain total quality score.In another example
A kind of above-mentioned convolutional neural networks of many indexes combined training can be directed to, which can directly export total matter
Amount scoring.
In step S230, face quality in the facial image for belonging to same target object is reached into preset requirement most earliest
Early facial image is sent to face recognition module and carries out recognition of face.
Target object can be anyone for including in image sequence.It, can be with during real-time perfoming Face datection
According to before current time the Face datection result of different images find out and belong to the face of same target, which is properly termed as
To image tracing.
It illustratively, can be in real time by the location information of each face detected from each image of image sequence
Input tracker module, tracker module by object tracking algorithm, can obtain it is relevant to each object at least one with
Track track, to know which face belongs to same target.For example, a large amount of band Trace Identifiers can be obtained by tracking
The bounding box of (track ID), the bounding box of same target can distribute identical track ID.That is, two are located at not
With the bounding box track ID having the same in image, this represents them and belongs to same target.It can be extracted based on bounding box
The facial image of face is corresponded to out, so also it is known which face or saying which facial image belongs to same target.Often
A track ID can represent a pursuit path, such as the track ID of object A can be 1, and the track ID of object B can be with
It is 2, and so on.Therefore, a pursuit path can be obtained for each object.
Assuming that before current time, from 10 image detections of image sequence to same target A, then can accordingly obtain
10 facial images of object A.It illustratively, can be by the people of the facial image when obtaining each facial image of object A
Face quality score is compared with quality threshold, to judge whether it reaches preset requirement.For example, quality threshold is 90 points, first 5
The face quality score of facial image is below the threshold value, but the face quality score of the 6th facial image is 91 points, then may be used
To determine that the 6th facial image is the earliest facial image that face quality reaches preset requirement earliest.At this point it is possible to by object A
The 6th facial image be sent to face recognition module carry out recognition of face.
Recognition of face may include comparing the face in facial image with reference to face, to identify facial image institute
Belong to the identity of object.The comparison of facial image and reference picture can be to be compared in pairs, is also possible to one-to-many comparison.Ability
Field technique personnel are understood that the implementation of recognition of face, do not repeat herein.
For currently detected any object, the face matter of each facial image of the object can be belonged to real-time judge
Whether amount reaches preset requirement.When find earliest reach the facial image of preset requirement after, which can be sent
Carry out recognition of face to face recognition module, though behind obtain the facial image for reaching preset requirement of the object again, also not
Face recognition module is sent it to again, unless it is the optimal facial image being buffered in optimal frame queue.
It is appreciated that above-mentioned send face recognition module progress recognition of face for the earliest facial image of target object
Operation executes in the presence of the earliest facial image of target object.During real-time image acquisition sequence, with
Time increases, and the picture number in image sequence is more and more, possibly can not to detect target pair in preceding several images
The face quality of elephant reaches preset requirement, but when getting some image, may detect that the face quality of target object
Reach preset requirement, can start to execute step S230 at this time.
In step S240, cache that face in the facial image for belonging to target object is best in quality and people in optimal frame queue
Face quality is more than the optimal facial image of earliest facial image.
Assuming that only sending earliest facial image carries out recognition of face, if error or even if face matter occurs in quality analysis
Amount, which reaches preset requirement, also can not correctly identify face, then may be such that the face of this object can not again identify.
Therefore, while carrying out recognition of face by earliest facial image, top-quality optimal facial image can also be cached, this
Sample can also be assisted carrying out face knowledge with optimal facial image in the case where that can not identify face by earliest facial image
Not.Calculation amount ratio needed for caching optimal facial image and carrying out these operations of auxiliary recognition of face by optimal facial image
Calculation amount needed for each facial image carries out recognition of face is much smaller, therefore combines earliest facial image and optimal face
Calculation amount needed for image carries out the scheme of recognition of face carries out the scheme of recognition of face much smaller than each facial image, and
The probability that its recognition of face goes out is higher than the identification probability that the scheme of single facial image is only considered for each object.It may be noted that
Herein, " recognition of face is carried out in conjunction with earliest facial image and optimal facial image " and referred to (such as earliest people when needed
When face image recognition failures) with optimal facial image recognition of face is assisted, and it is not necessarily referring to that earliest facial image must be utilized simultaneously
Recognition of face is carried out with optimal facial image.
Illustratively, optimal frame queue can store in memory.Optimal frame queue can take up certain memory headroom,
In optimal frame queue, storage allocation space can be distinguished for each object, for caching the optimal facial image of the object.It is right
In same target, optimal frame queue can only cache face quality so far and reach preset requirement and the top-quality list of face
A facial image.
It is appreciated that the above-mentioned operation for caching optimal facial image in optimal frame queue executes after step S230.
Until at the time of obtaining earliest facial image, earliest facial image is the top-quality facial image of face, later with figure
As the picture number in sequence is more and more, it is possible to detect that face quality is more than the facial image of earliest facial image, this
When the facial image can be buffered in optimal frame queue.If occurring the optimal facial image people than currently having cached below
The facial image of face better quality can then use the facial image to be cached as new optimal facial image.
In step S250, at the scheduled time, the image buffer storage situation of optimal frame queue is checked, if delayed in optimal frame queue
The face for having the optimal facial image for belonging to special object and special object not yet identifies, then will belong to special object most
Excellent facial image is sent to face recognition module and carries out recognition of face.
Although being carried out as described above, sending face recognition module for the earliest facial image for belonging to certain an object
Recognition of face, it is possible that success can not be identified.For this class object, auxiliary knowledge can be carried out by its optimal facial image
Not.
At the time of predetermined instant can be any appropriate.For example, predetermined instant, which can be target object, passes through its earliest people
Face image carries out a certain moment after recognition of face failure.In another example predetermined instant can also be the self clock moment (under such as
At the time of stating opening timing device) start at the time of predetermined period is reached.
For example, can just detect in optimal frame queue whether be cached with facial image every predetermined period, if so,
It may determine that whether the face of the affiliated object of the facial image has been identified, if identified, can incite somebody to action
The facial image is stored in nonvolatile memory (such as disk), is stored for a long time, can be with if not yet identified
Recognition of face is carried out using the facial image.
Target object involved in step S230 and step S240 can be any object.It in one example, can be with
Step S230 and S240 are executed respectively to each object detected from image sequence, that is, detect from image sequence
Each object step S230 and S240 can be executed respectively as target object.In another example, can only to from
Each object in the partial objects detected in image sequence executes step S230 and S240 respectively.For example, can be by user
It specifies or system automatically selects some or multiple objects execute step S230 and S240 respectively as target object.
Special object involved in step S250 can be any object.In one example, the special object is
Target object involved in step S230 and step S240, such as target object can not successfully identified by step S230
In the case where face, its optimal facial image can be read from optimal frame queue and carries out recognition of face.In this case, it examines
When looking into the image buffer storage situation of optimal frame queue, specific aim inspection is carried out, i.e., only checks whether the face for being cached with target object
Image.In another example, the special object can be optimal frame queue at the scheduled time and be cached with its facial image
Any object.In this case, when checking the image buffer storage situation of optimal frame queue, non-specific aim inspection is carried out, that is, is examined
The facial image for whether being cached with any object is looked into, it, can be by this if being cached with the facial image of one or more objects
A or multiple objects are regarded as special object respectively.It is understood that in latter case, if at the scheduled time, executed
Step S230 and S240 are crossed, i.e., if having cached its optimal facial image for the target object, then non-specific aim checks
The facial image of target object can be checked, target object can be considered special object at this time.
It may be noted that each step of image processing method 200 shown in Fig. 2 is not limited to sequence executive mode, for example, step
Rapid S210 and S220 is executed in real time, and step S210 can in real time, in turn obtain the figure one by one in image sequence
Picture, one image of every acquisition can carry out Face datection involved in step S220, facial image extraction and face to it
Quality analysis and etc..In addition, step S230, S240 and S250 have the respective execution time respectively or execute condition, this can be with
Understand in conjunction with being described herein.
Image processing method according to an embodiment of the present invention can select face quality earliest qualified (reaching preset requirement)
Most fast facial image carry out recognition of face, rather than each facial image carries out recognition of face, can effectively reduce nothing in this way
The repetition of meaning identifies, that is, reduces useless calculating, while additionally aiding and rapidly identifying face as far as possible.At the same time, the party
Method is also buffered in the top-quality optimal facial image of face after earliest facial image, when someone passes through its earliest face
When image can not successfully identify face, recognition of face can be carried out with the optimal facial image of caching, people can be improved in this way
The probability that face is identified.As it can be seen that the recognition of face strategy that combines of this earliest facial image and optimal facial image can be with
It effectively reduces useless calculating, identify face as early as possible, while the probability that recognition of face goes out can be improved as far as possible.
Illustratively, image processing method according to an embodiment of the present invention can be in setting with memory and processor
It is realized in standby, device or system.
Image processing method according to an embodiment of the present invention can be deployed at Image Acquisition end, for example, in security protection application
Field can be deployed in the Image Acquisition end of access control system；In financial application field, can be deployed at personal terminal, such as
Smart phone, tablet computer, personal computer etc..
Alternatively, image processing method according to an embodiment of the present invention can also be deployed in server end (or cloud with being distributed
End) and personal terminal at.For example, can client obtain image, the image that client will acquire send to server end (or
Cloud), image procossing is carried out by server end (or cloud).
According to embodiments of the present invention, face quality in the facial image for belonging to same target object is reached to default earliest to want
It may include: that real-time judge belongs to that the earliest facial image asked, which is sent to face recognition module and carries out recognition of face (step S230),
Whether the face quality of each facial image reaches preset requirement in the facial image of target object, until earliest facial image goes out
It is existing；And face recognition module is sent by earliest facial image and carries out recognition of face.
After the earliest facial image of target object occurs, it may not need and judge whether subsequent facial image reaches again
Preset requirement can reduce calculation amount in this way.
According to embodiments of the present invention, face quality in the facial image for belonging to same target object is reached to default earliest to want
It may include: that real-time judge belongs to that the earliest facial image asked, which is sent to face recognition module and carries out recognition of face (step S230),
Whether the face quality of each facial image reaches preset requirement in the facial image of target object；In the people for belonging to target object
In the case that the face quality of current face's image in face image reaches preset requirement, judgement belongs to the face figure of target object
It is already sent to face recognition module with the presence or absence of any Prior face's image as in, if there is no then determining current face's figure
It seem earliest facial image；Face recognition module, which is sent, by earliest facial image carries out recognition of face.
Whether its face quality, which reaches preset requirement, is judged for each facial image of target object, is reached if found
To the facial image of preset requirement, it is also necessary to which whether other images that existing face quality reaches preset requirement are sent out before for judgement
It is sent to face recognition module, does not retransmit current face's image if it is, otherwise can send people for current face's image
Face identification module carries out recognition of face.
Illustratively, it can be each object allocation identification information detected, be used to indicate the object with the presence or absence of logical
It crosses step S230 and is judged as that face quality reaches the facial image of preset requirement and is sent to face recognition module.For example, target
The identification information of object can be a flag bit (flag), when finding earliest facial image and send the facial image to
After face recognition module, the data of the flag bit can be set to 1 from 0, the subsequent new face for getting target object again
When image, it can know that existing Prior face's image is sent to face recognition module by reading flag bit, at this time may not be used
Retransmit the new facial image.
According to embodiments of the present invention, in optimal frame queue cache belong to target object facial image in face quality most
Good and face quality is more than that the optimal facial image (step S240) of earliest facial image may include: to belong to target object
It is more than the facial image of the face quality of earliest facial image as optimal face figure using first man face quality in facial image
As being buffered in optimal frame queue；When acquisition belongs to current face's image of target object every time, by the people of current face's image
Face quality and the face quality of optimal facial image compare；It is more than optimal face figure in the face quality of current face's image
In the case where the face quality of picture, the optimal facial image cached in optimal frame queue is updated to current face's image.
For example, for certain an object A its earliest facial image I can be found by step S230 firsta1.Then,
In the next man's face image I for obtaining object Aa2It later, can be by Ia2Face quality and Ia1Face quality compare,
If Ia2Face better quality, then can be by Ia2It is buffered in optimal frame queue.Then, each subsequent of object A is being obtained
Facial image IaiLater, every time by IaiFace quality versus is carried out with the facial image of current cache, if IaiFace matter
Amount is more preferable to be then cached the facial image as new optimal facial image.It is appreciated that optimal facial image is face
Quality reaches preset requirement and face quality is higher than earliest facial image.In this way, it is ensured that optimal frame queue
In cache object A always the top-quality facial image of face.When Finding Object A at the scheduled time is not yet identified, then may be used
To carry out recognition of face using the facial image of the object A of caching.
Above scheme can make optimal frames queue remain caching mesh by carrying out real-time update to optimal frame queue
The top-quality facial image of face of object is marked, the face quality of the facial image is more than the earliest facial image of target object
Face quality, help to improve recognition of face go out probability.
According to embodiments of the present invention, target object and special object are same targets, at the scheduled time, check optimal frames team
The image buffer storage situation (step S250) of column may include: the case where carrying out recognition of face failure based on earliest facial image
Under, check the facial image for whether being cached in optimal frame queue and belonging to target object, wherein predetermined instant is based on earliest people
Face image carries out any time after recognition of face failure.
As described above, checking the image buffer storage of optimal frame queue it may is that specific aim inspection.That is, passing through step
In the case that S230 can not successfully identify the face of target object, can be read from optimal frame queue its optimal facial image into
Row recognition of face.This scheme can select the object for carrying out auxiliary recognition of face from face recognition result one end, avoid
The operations such as caching inspection are carried out to the successful object of recognition of face, calculation amount can be reduced to a certain extent.
According to embodiments of the present invention, at the scheduled time, check that the image buffer storage situation (step S250) of optimal frame queue can
To include: to check the facial image for whether being cached in optimal frame queue and belonging to any object every predetermined period, wherein special
Determining object is any object that facial image is buffered in optimal frame queue.
It is appreciated that in the present embodiment, predetermined instant can be the self clock moment and start timing, every predetermined period institute
At the time of arrival.Predetermined period can be any suitable period, and the present invention limits not to this.Illustratively, pre- timing
Section can be 2 seconds, 5 seconds, 10 seconds etc..Illustratively, the timer with the fixed cycle can be set in face identification system,
The period is the predetermined period, the timer can using hardware, software or its be implemented in combination with.Assuming that the week of timer
Phase is 2 seconds, then predetermined instant can be the 2nd second, the 4th second, the 6th second ... of the timing since the opening timing device.
As described above, checking the image buffer storage of optimal frame queue it may is that non-specific aim inspection.This non-specific aim
Inspection optionally can be executed periodically, such as be checked once every execution in 2 seconds.If current check is into optimal frame queue one
The optimal facial image of two object C and object D have been cached altogether, then can judge the object for object C and object D respectively
Whether face has identified, if any one object does not identify, can use the optimal facial image of the object
Carry out recognition of face.This scheme can select pair for carrying out auxiliary recognition of face from the facial image one end cached
As may not need consideration for the object of no caching facial image and carry out auxiliary recognition of face, the work of system can be improved
Efficiency.In addition, inspecting periodically the image buffer storage situation of optimal frame queue, the object that facial image has cached can be arranged in time,
Auxiliary recognition of face (when it is needed) is carried out to these objects as early as possible, and then recognition of face efficiency can be improved.
Above-described embodiment is only example, non-specific aim inspection can also some or it is multiple set at the time of execution,
Or the variable cycle can be based on rather than fixed cycle execution.
According to embodiments of the present invention, in the image buffer storage situation (step S250) at the scheduled time, checking optimal frame queue
Later, if method 200 can also include: to be cached with the optimal facial image for belonging to special object and specific in optimal frame queue
The face of object has identified, then the optimal facial image for belonging to special object is removed from optimal frame queue and/or will be belonged to
The optimal facial image storage of special object is in memory.
Memory space can be saved by removing the optimal facial image for having participated in identifying in time, reduce further cache inspection pressure
Power, and then the working efficiency of face identification system can be improved.Optimal facial image is stored in such as above-mentioned non-volatile deposit
In the memory of reservoir, other operations can be executed by the facial image in order to subsequent calls, such as be displayed for over the display
User checks that the same day participates in the personnel etc. of recognition of face.
According to embodiments of the present invention, Face datection (step S220) is carried out to each image in image sequence in real time
When, method 200 can also include: two faces for being located in the two of image sequence different images, based on this two
Whether the diversity judgement between location information two faces in the Face datection result of a face belong to same target object.
As set forth above, it is possible to which the location information based on face is carried out to image tracing, to judge which face belongs to a pair
As.Position difference of the face of same target between former and later two images is not too large, therefore, can be by former and later two
Difference in image between the face location of two faces judges whether they belong to same target.Above-mentioned judgement belongs to same
The mode of the face of object can be adapted for any object.For example, for arbitrary two different images for being located at image sequence
In two faces, can diversity judgement this two between the location information in the Face datection result based on two faces
Whether face belongs to same target.
Determine whether the face in two images is belonged to same target and can be likely to occur using any existing or future
Object tracking mode realize, it is not limited to example described herein.
Fig. 3 shows the schematic diagram of an exemplary image processing flow according to the present invention.As shown in figure 3, firstly, opening phase
Machine starts to acquire image, while can open the timer that the period is 2 seconds.Then, one image of every acquisition just carries out face inspection
It surveys, if detecting face, the image block where each face can be extracted, obtain facial image, and analyze face figure
The face quality of picture.For certain an object, it can be determined that whether the face quality of each of which facial image is qualified (to reach
Preset requirement), it can also judge whether this person has been participating in identification in the case where qualification, i.e., previously whether have face figure
As being sent to face recognition module, if it is not, can send current face's image to face recognition module carries out people
Face identification.In addition, also carry out facial image choose it is excellent.It can be more than first earliest by first man face quality for certain an object
In the facial image caching optimal frame queue in memory of the face quality of facial image, behind if any current face schemes
As the face better quality of the optimal facial image than caching, then by current face's image buffer storage into optimal frame queue.Every time 2
When the period of second timer reaches, detect and whether be cached with facial image (i.e. data) in optimal frame queue, if there is and at this time
It not yet identifies the face of the affiliated object of the facial image, then can send recognition of face for the optimal facial image of the object
Module carries out recognition of face can deposit optimal facial image if having identified the face of the affiliated object of the facial image
Store up disk.
According to a further aspect of the invention, a kind of image processing apparatus is provided.Fig. 4 is shown according to an embodiment of the present invention
Image processing apparatus 400 schematic block diagram.
As shown in figure 4, image processing apparatus 400 according to an embodiment of the present invention includes obtaining module 410, detection and analysis
Module 420, sending module 430, cache module 440 and inspection module 450.The modules can execute respectively above in conjunction with
The each step/function for the image processing method that Fig. 2-3 is described.Below only to the master of each component of the image processing apparatus 400
It wants function to be described, and omits the detail content having been described above.
It obtains module 410 and is used for real-time image acquisition sequence.Obtaining module 410 can be in electronic equipment as shown in Figure 1
102 Running storage device 104 of processor in the program instruction that stores realize.
Detection and analysis module 420 is being examined for carrying out Face datection to each image in described image sequence in real time
Facial image corresponding to each face is obtained in the case where measuring face, and analyzes the face quality of each facial image.Inspection
It surveys and journey that analysis module 420 can store in 102 Running storage device 104 of processor in electronic equipment as shown in Figure 1
Sequence instructs to realize.
Sending module 430 is used to face quality in the facial image for belonging to same target object reaching preset requirement earliest
Earliest facial image be sent to face recognition module carry out recognition of face.Sending module 430 can electronics as shown in Figure 1 set
The program instruction that stores in 102 Running storage device 104 of processor in standby is realized.
Cache module 440 in optimal frame queue for caching face quality in the facial image for belonging to the target object
Best and face quality is more than the optimal facial image of the earliest facial image.Cache module 440 can electricity as shown in Figure 1
The program instruction that stores in 102 Running storage device 104 of processor in sub- equipment is realized.
Check that module 450 at the scheduled time, checks the image buffer storage situation of the optimal frame queue, if it is described most
The face that the optimal facial image and the special object that belong to special object are cached in excellent frame queue not yet identifies, then will
The optimal facial image for belonging to the special object is sent to the face recognition module and carries out recognition of face.Check module 450
The program instruction that can store in 102 Running storage device 104 of processor in electronic equipment as shown in Figure 1 is realized.
Those of ordinary skill in the art may be aware that list described in conjunction with the examples disclosed in the embodiments of the present disclosure
Member and algorithm steps can be realized with the combination of electronic hardware or computer software and electronic hardware.These functions are actually
It is implemented in hardware or software, the specific application and design constraint depending on technical solution.Professional technician
Each specific application can be used different methods to achieve the described function, but this realization is it is not considered that exceed
The scope of the present invention.
Fig. 5 shows the schematic block diagram of image processing system 500 according to an embodiment of the invention.Image procossing system
System 500 includes image collecting device 510, storage device (i.e. memory) 520 and processor 530.
Described image acquisition device 510 is for acquiring image.Image collecting device 510 is optional, image processing system
500 can not include image collecting device 510.In such a case, it is possible to using other image acquisition device images, and
The image of acquisition is sent to image processing system 500.
The storage of storage device 520 is for realizing the corresponding steps in image processing method according to an embodiment of the present invention
Computer program instructions.
The processor 530 is for running the computer program instructions stored in the storage device 520, to execute basis
The corresponding steps of the image processing method of the embodiment of the present invention.
In one embodiment, for executing following step when the computer program instructions are run by the processor 530
It is rapid: real-time image acquisition sequence；Face datection is carried out to each image in image sequence in real time, the case where detecting face
It is lower to obtain facial image corresponding to each face, and analyze the face quality of each facial image；Same target pair will be belonged to
The earliest facial image that face quality reaches preset requirement earliest in the facial image of elephant is sent to face recognition module and carries out people
Face identification；Face is best in quality in the facial image that caching belongs to target object in optimal frame queue and face quality is more than most
The optimal facial image of early facial image；At the scheduled time, the image buffer storage situation of optimal frame queue is checked, if optimal frames team
The face that the optimal facial image and special object that belong to special object are cached in column not yet identifies that it is specific right then to belong to
The optimal facial image of elephant is sent to face recognition module and carries out recognition of face.
In addition, according to embodiments of the present invention, additionally providing a kind of storage medium, storing program on said storage
Instruction, when described program instruction is run by computer or processor for executing the image processing method of the embodiment of the present invention
Corresponding steps, and for realizing the corresponding module in image processing apparatus according to an embodiment of the present invention.The storage medium
It such as may include the storage card of smart phone, the storage unit of tablet computer, the hard disk of personal computer, read-only memory
(ROM), Erasable Programmable Read Only Memory EPROM (EPROM), portable compact disc read-only memory (CD-ROM), USB storage,
Or any combination of above-mentioned storage medium.
In one embodiment, described program instruction can make computer or place when being run by computer or processor
Reason device realizes each functional module of image processing apparatus according to an embodiment of the present invention, and and/or can execute according to this
The image processing method of inventive embodiments.
In one embodiment, described program instruction is at runtime for executing following steps: real-time image acquisition sequence；
Face datection is carried out to each image in image sequence in real time, is obtained in the case where detecting face corresponding to each face
Facial image, and analyze the face quality of each facial image；By face matter in the facial image for belonging to same target object
The earliest facial image that amount reaches preset requirement earliest is sent to face recognition module and carries out recognition of face；In optimal frame queue
Caching belongs to that face in the facial image of target object is best in quality and face quality is more than the optimal face of earliest facial image
Image；At the scheduled time, check the image buffer storage situation of optimal frame queue, if be cached in optimal frame queue belong to it is specific right
The optimal facial image of the elephant and face of special object not yet identifies, the then optimal facial image that will belong to special object are sent
Recognition of face is carried out to face recognition module.
Each module in image processing system according to an embodiment of the present invention can pass through reality according to an embodiment of the present invention
The processor computer program instructions that store in memory of operation of the electronic equipment of image procossing are applied to realize, or can be with
The computer instruction stored in the computer readable storage medium of computer program product according to an embodiment of the present invention is counted
Calculation machine is realized when running.
Although describing example embodiment by reference to attached drawing here, it should be understood that above example embodiment are only exemplary
, and be not intended to limit the scope of the invention to this.Those of ordinary skill in the art can carry out various changes wherein
And modification, it is made without departing from the scope of the present invention and spiritual.All such changes and modifications are intended to be included in appended claims
Within required the scope of the present invention.
Those of ordinary skill in the art may be aware that list described in conjunction with the examples disclosed in the embodiments of the present disclosure
Member and algorithm steps can be realized with the combination of electronic hardware or computer software and electronic hardware.These functions are actually
It is implemented in hardware or software, the specific application and design constraint depending on technical solution.Professional technician
Each specific application can be used different methods to achieve the described function, but this realization is it is not considered that exceed
The scope of the present invention.
In several embodiments provided herein, it should be understood that disclosed device and method can pass through it
Its mode is realized.For example, apparatus embodiments described above are merely indicative, for example, the division of the unit, only
Only a kind of logical function partition, there may be another division manner in actual implementation, such as multiple units or components can be tied
Another equipment is closed or is desirably integrated into, or some features can be ignored or not executed.
In the instructions provided here, numerous specific details are set forth.It is to be appreciated, however, that implementation of the invention
Example can be practiced without these specific details.In some instances, well known method, structure is not been shown in detail
And technology, so as not to obscure the understanding of this specification.
Similarly, it should be understood that in order to simplify the present invention and help to understand one or more of the various inventive aspects,
To in the description of exemplary embodiment of the present invention, each feature of the invention be grouped together into sometimes single embodiment, figure,
Or in descriptions thereof.However, the method for the invention should not be construed to reflect an intention that i.e. claimed
The present invention claims features more more than feature expressly recited in each claim.More precisely, such as corresponding power
As sharp claim reflects, inventive point is that the spy of all features less than some disclosed single embodiment can be used
Sign is to solve corresponding technical problem.Therefore, it then follows thus claims of specific embodiment are expressly incorporated in this specific
Embodiment, wherein each, the claims themselves are regarded as separate embodiments of the invention.
It will be understood to those skilled in the art that any combination pair can be used other than mutually exclusive between feature
All features disclosed in this specification (including adjoint claim, abstract and attached drawing) and so disclosed any method
Or all process or units of equipment are combined.Unless expressly stated otherwise, this specification (is wanted including adjoint right
Ask, make a summary and attached drawing) disclosed in each feature can be replaced with an alternative feature that provides the same, equivalent, or similar purpose.
In addition, it will be appreciated by those of skill in the art that although some embodiments described herein include other embodiments
In included certain features rather than other feature, but the combination of the feature of different embodiments mean it is of the invention
Within the scope of and form different embodiments.For example, in detail in the claims, embodiment claimed it is one of any
Can in any combination mode come using.
Various component embodiments of the invention can be implemented in hardware, or to run on one or more processors
Software module realize, or be implemented in a combination thereof.It will be understood by those of skill in the art that can be used in practice
Microprocessor or digital signal processor (DSP) realize some moulds in image processing apparatus according to an embodiment of the present invention
The some or all functions of block.The present invention is also implemented as a part or complete for executing method as described herein
The program of device (for example, computer program and computer program product) in portion.It is such to realize that program of the invention can store
On a computer-readable medium, it or may be in the form of one or more signals.Such signal can be from internet
Downloading obtains on website, is perhaps provided on the carrier signal or is provided in any other form.
It should be noted that the above-mentioned embodiments illustrate rather than limit the invention, and ability
Field technique personnel can be designed alternative embodiment without departing from the scope of the appended claims.In the claims,
Any reference symbol between parentheses should not be configured to limitations on claims.Word "comprising" does not exclude the presence of not
Element or step listed in the claims.Word "a" or "an" located in front of the element does not exclude the presence of multiple such
Element.The present invention can be by means of including the hardware of several different elements and being come by means of properly programmed computer real
It is existing.In the unit claims listing several devices, several in these devices can be through the same hardware branch
To embody.The use of word first, second, and third does not indicate any sequence.These words can be explained and be run after fame
Claim.
The above description is merely a specific embodiment or to the explanation of specific embodiment, protection of the invention
Range is not limited thereto, and anyone skilled in the art in the technical scope disclosed by the present invention, can be easily
Expect change or replacement, should be covered by the protection scope of the present invention.Protection scope of the present invention should be with claim
Subject to protection scope.
Claims (11)
1. a kind of image processing method, comprising:
Real-time image acquisition sequence；
Face datection is carried out to each image in described image sequence in real time, obtains everyone in the case where detecting face
Facial image corresponding to face, and analyze the face quality of each facial image；
The earliest facial image that face quality in the facial image for belonging to same target object reaches preset requirement earliest is sent
Recognition of face is carried out to face recognition module；
Caching belongs to that face is best in quality in the facial image of the target object and face quality is more than in optimal frame queue
The optimal facial image of the earliest facial image；
At the scheduled time, the image buffer storage situation for checking the optimal frame queue, if being cached with category in the optimal frame queue
In the optimal facial image of special object and the face of the special object not yet identifies, then will belong to the special object
Optimal facial image is sent to the face recognition module and carries out recognition of face.
2. the method for claim 1, wherein it is described by face quality in the facial image for belonging to same target object most
The earliest facial image for early reaching preset requirement is sent to face recognition module progress recognition of face and includes:
Whether the face quality that real-time judge belongs to each facial image in the facial image of the target object reaches described pre-
If it is required that until the earliest facial image occurs；And
The face recognition module, which is sent, by the earliest facial image carries out recognition of face.
3. the method for claim 1, wherein it is described by face quality in the facial image for belonging to same target object most
The earliest facial image for early reaching preset requirement is sent to face recognition module progress recognition of face and includes:
Whether the face quality that real-time judge belongs to each facial image in the facial image of the target object reaches described pre-
If it is required that；
The face quality of current face's image in the facial image for belonging to the target object reaches the preset requirement
In the case of, judgement, which belongs in the facial image of the target object, is already sent to face with the presence or absence of any Prior face's image
Identification module, if there is no then determining that current face's image is the earliest facial image；
The face recognition module, which is sent, by the earliest facial image carries out recognition of face.
4. the method for claim 1, wherein the caching in optimal frame queue belongs to the face of the target object
Face is best in quality in image and face quality is more than that the optimal facial image of the earliest facial image includes:
It is more than the face of the earliest facial image by first man face quality in the facial image for belonging to the target object
The facial image of quality is buffered in the optimal frame queue as the optimal facial image；
When acquisition belongs to current face's image of the target object every time, by the face quality of current face's image and institute
The face quality for stating optimal facial image compares；
It, will be described in the case where the face quality of current face's image is more than the face quality of the optimal facial image
The optimal facial image cached in optimal frame queue is updated to current face's image.
5. the method for claim 1, wherein the target object and the special object are same targets, it is described
Predetermined instant checks that the image buffer storage situation of the optimal frame queue includes:
In the case where carrying out recognition of face failure based on the earliest facial image, check whether delay in the optimal frame queue
There is the facial image for belonging to the target object, wherein the predetermined instant is to carry out people based on the earliest facial image
Any time after face recognition failures.
6. described in the method for claim 1, wherein at the scheduled time, checking the image buffer storage of the optimal frame queue
Situation includes:
Every predetermined period, check in the optimal frame queue whether be cached with the facial image for belonging to any object, wherein institute
Stating special object is any object that facial image is buffered in the optimal frame queue.
7. the method for claim 1, wherein it is described at the scheduled time, check that the image of the optimal frame queue is slow
After depositing situation, the method also includes:
If being cached with the optimal facial image for belonging to the special object and the special object in the optimal frame queue
Face has identified, then the optimal facial image for belonging to the special object is removed from the optimal frame queue and/or will be belonged to
It is stored in memory in the optimal facial image of the special object.
8. the method for claim 1, wherein carrying out face to each image in described image sequence in real time described
When detection, the method also includes:
For two faces being located in two different images of described image sequence, the face inspection based on two faces
Whether diversity judgement two faces between location information surveyed in result belong to same target object.
9. a kind of image processing apparatus, comprising:
Module is obtained, real-time image acquisition sequence is used for；
Detection and analysis module is detecting people for carrying out Face datection to each image in described image sequence in real time
Facial image corresponding to each face is obtained in the case where face, and analyzes the face quality of each facial image；
Sending module, for face quality in the facial image for belonging to same target object to be reached the earliest of preset requirement earliest
Facial image is sent to face recognition module and carries out recognition of face；
Cache module, in optimal frame queue caching belong to face in the facial image of the target object it is best in quality and
Face quality is more than the optimal facial image of the earliest facial image；
Module is checked, at the scheduled time, the image buffer storage situation of the optimal frame queue being checked, if the optimal frames team
The face that the optimal facial image and the special object that belong to special object are cached in column not yet identifies, then will belong to institute
The optimal facial image for stating special object is sent to the face recognition module and carries out recognition of face.
10. a kind of image processing system, including processor and memory, wherein be stored with computer program in the memory
Instruction, for executing figure as claimed in any one of claims 1 to 8 when the computer program instructions are run by the processor
As processing method.
11. a kind of storage medium stores program instruction on said storage, described program instruction is at runtime for holding
Row image processing method as claimed in any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910477796.2A CN110263680B (en) | 2019-06-03 | 2019-06-03 | Image processing method, device and system and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910477796.2A CN110263680B (en) | 2019-06-03 | 2019-06-03 | Image processing method, device and system and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110263680A true CN110263680A (en) | 2019-09-20 |
| CN110263680B CN110263680B (en) | 2022-01-28 |
Family
ID=67916525
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910477796.2A Active CN110263680B (en) | 2019-06-03 | 2019-06-03 | Image processing method, device and system and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110263680B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110852303A (en) * | 2019-11-21 | 2020-02-28 | 中科智云科技有限公司 | Eating behavior identification method based on OpenPose |
| CN111160221A (en) * | 2019-12-26 | 2020-05-15 | 深圳云天励飞技术有限公司 | Face acquisition method and related device |
| CN111783512A (en) * | 2019-11-11 | 2020-10-16 | 西安宇视信息科技有限公司 | Image processing method, device, equipment and storage medium |
| CN111860163A (en) * | 2020-06-17 | 2020-10-30 | 北京百度网讯科技有限公司 | Image processing method and device, electronic equipment and readable storage medium |
| WO2021208250A1 (en) * | 2020-04-15 | 2021-10-21 | 上海摩象网络科技有限公司 | Face tracking method and face tracking device |
| CN111783512B (en) * | 2019-11-11 | 2024-05-14 | 西安宇视信息科技有限公司 | Image processing method, device, equipment and storage medium |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2007274264A (en) * | 2006-03-31 | 2007-10-18 | Casio Comput Co Ltd | Camera, photographing method of best-shot image and program |
| CN101458763A (en) * | 2008-10-30 | 2009-06-17 | 中国人民解放军国防科学技术大学 | Automatic human face identification method based on image weighting average |
| US20140240453A1 (en) * | 2013-02-26 | 2014-08-28 | Samsung Electronics Co., Ltd. | Apparatus and method photographing image |
| CN104978550A (en) * | 2014-04-08 | 2015-10-14 | 上海骏聿数码科技有限公司 | Face recognition method and system based on large-scale face database |
| CN105938552A (en) * | 2016-06-29 | 2016-09-14 | 北京旷视科技有限公司 | Face recognition method capable of realizing base image automatic update and face recognition device |
| CN107273862A (en) * | 2017-06-20 | 2017-10-20 | 深圳市乐易时代科技有限公司 | A kind of automatic grasp shoot method, monitoring device and computer-readable recording medium |
| CN107346426A (en) * | 2017-07-10 | 2017-11-14 | 深圳市海清视讯科技有限公司 | A kind of face information collection method based on video camera recognition of face |
| CN108491784A (en) * | 2018-03-16 | 2018-09-04 | 南京邮电大学 | The identification in real time of single feature towards large-scale live scene and automatic screenshot method |
| CN108875512A (en) * | 2017-12-05 | 2018-11-23 | 北京旷视科技有限公司 | Face identification method, device, system, storage medium and electronic equipment |
| CN108875452A (en) * | 2017-05-11 | 2018-11-23 | 北京旷视科技有限公司 | Face identification method, device, system and computer-readable medium |
| CN109034013A (en) * | 2018-07-10 | 2018-12-18 | 腾讯科技（深圳）有限公司 | A kind of facial image recognition method, device and storage medium |
| CN109035246A (en) * | 2018-08-22 | 2018-12-18 | 浙江大华技术股份有限公司 | A kind of image-selecting method and device of face |
-
2019
- 2019-06-03 CN CN201910477796.2A patent/CN110263680B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2007274264A (en) * | 2006-03-31 | 2007-10-18 | Casio Comput Co Ltd | Camera, photographing method of best-shot image and program |
| CN101458763A (en) * | 2008-10-30 | 2009-06-17 | 中国人民解放军国防科学技术大学 | Automatic human face identification method based on image weighting average |
| US20140240453A1 (en) * | 2013-02-26 | 2014-08-28 | Samsung Electronics Co., Ltd. | Apparatus and method photographing image |
| CN104978550A (en) * | 2014-04-08 | 2015-10-14 | 上海骏聿数码科技有限公司 | Face recognition method and system based on large-scale face database |
| CN105938552A (en) * | 2016-06-29 | 2016-09-14 | 北京旷视科技有限公司 | Face recognition method capable of realizing base image automatic update and face recognition device |
| CN108875452A (en) * | 2017-05-11 | 2018-11-23 | 北京旷视科技有限公司 | Face identification method, device, system and computer-readable medium |
| CN107273862A (en) * | 2017-06-20 | 2017-10-20 | 深圳市乐易时代科技有限公司 | A kind of automatic grasp shoot method, monitoring device and computer-readable recording medium |
| CN107346426A (en) * | 2017-07-10 | 2017-11-14 | 深圳市海清视讯科技有限公司 | A kind of face information collection method based on video camera recognition of face |
| CN108875512A (en) * | 2017-12-05 | 2018-11-23 | 北京旷视科技有限公司 | Face identification method, device, system, storage medium and electronic equipment |
| CN108491784A (en) * | 2018-03-16 | 2018-09-04 | 南京邮电大学 | The identification in real time of single feature towards large-scale live scene and automatic screenshot method |
| CN109034013A (en) * | 2018-07-10 | 2018-12-18 | 腾讯科技（深圳）有限公司 | A kind of facial image recognition method, device and storage medium |
| CN109035246A (en) * | 2018-08-22 | 2018-12-18 | 浙江大华技术股份有限公司 | A kind of image-selecting method and device of face |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111783512A (en) * | 2019-11-11 | 2020-10-16 | 西安宇视信息科技有限公司 | Image processing method, device, equipment and storage medium |
| CN111783512B (en) * | 2019-11-11 | 2024-05-14 | 西安宇视信息科技有限公司 | Image processing method, device, equipment and storage medium |
| CN110852303A (en) * | 2019-11-21 | 2020-02-28 | 中科智云科技有限公司 | Eating behavior identification method based on OpenPose |
| CN111160221A (en) * | 2019-12-26 | 2020-05-15 | 深圳云天励飞技术有限公司 | Face acquisition method and related device |
| CN111160221B (en) * | 2019-12-26 | 2023-09-01 | 深圳云天励飞技术有限公司 | Face acquisition method and related device |
| WO2021208250A1 (en) * | 2020-04-15 | 2021-10-21 | 上海摩象网络科技有限公司 | Face tracking method and face tracking device |
| CN111860163A (en) * | 2020-06-17 | 2020-10-30 | 北京百度网讯科技有限公司 | Image processing method and device, electronic equipment and readable storage medium |
| CN111860163B (en) * | 2020-06-17 | 2023-08-22 | 北京百度网讯科技有限公司 | Image processing method, device, electronic equipment and readable storage medium |
| US11783565B2 (en) | 2020-06-17 | 2023-10-10 | Beijing Baidu Netcom Science And Technology Co., Ltd. | Image processing method, electronic device and readable storage medium for maintaining a face image buffer queue |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110263680B (en) | 2022-01-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110263680A (en) | Image processing method, device and system and storage medium | |
| US11195037B2 (en) | Living body detection method and system, computer-readable storage medium | |
| CN109359548B (en) | Multi-face recognition monitoring method and device, electronic equipment and storage medium | |
| US11055516B2 (en) | Behavior prediction method, behavior prediction system, and non-transitory recording medium | |
| US9001199B2 (en) | System and method for human detection and counting using background modeling, HOG and Haar features | |
| CN108875522A (en) | Face cluster methods, devices and systems and storage medium | |
| CN108875452A (en) | Face identification method, device, system and computer-readable medium | |
| CN106407914A (en) | Method for detecting human faces, device and remote teller machine system | |
| CN108898104A (en) | A kind of item identification method, device, system and computer storage medium | |
| CN105518711A (en) | In-vivo detection method, in-vivo detection system, and computer program product | |
| CN108875546A (en) | Face auth method, system and storage medium | |
| CN105518708A (en) | Method and equipment for verifying living human face, and computer program product | |
| CN103279744B (en) | Imitation fingerprint detection methods based on multiple dimensioned three mode texture feature and system | |
| CN109815960A (en) | Reproduction image-recognizing method, device, equipment and medium based on deep learning | |
| CN108875517A (en) | Method for processing video frequency, device and system and storage medium | |
| CN112884092A (en) | AI model generation method, electronic device, and storage medium | |
| CN106650670A (en) | Method and device for detection of living body face video | |
| CN109255256A (en) | Distributed attendance record storage method, device and network node | |
| CN109460733A (en) | Recognition of face in-vivo detection method and system based on single camera, storage medium | |
| CN111652087A (en) | Car checking method and device, electronic equipment and storage medium | |
| CN106203024A (en) | Method and system are signed in a kind of long-distance video face | |
| CN108875509A (en) | Biopsy method, device and system and storage medium | |
| CN108875476A (en) | Automatic near-infrared face registration and recognition methods, device and system and storage medium | |
| KR20150128510A (en) | Apparatus and method for liveness test, and apparatus and method for image processing | |
| CN112153373A (en) | Fault identification method and device for bright kitchen range equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |